This project aims to elucidate the molecular nature of differences in murine leukemia virus (MuLV) infective host range in vitro, and pathogenicity in vivo, in the hope that characterization of the determinants responsible for the restriction of retroviral replication will prove generally enlightening. As a model system, we are examining an ecotropic MuLV (named RF.AMLV) which, along with the Moloney ecotropic MuLV (MoMuLV eco), is greatly restricted for in vitro infection of a Mus dunni mouse fibroblast cell line. This is quite unusual, since the Mus dunni cell line is highly sensitive to infection by most ecotropic MuLV and indeed to many xenotropic, amphotropic and mink cell focus-inducing (MCF) MuLV. Successive serial passage of the RF.AMLV through the dunni cell line gave rise to virus stocks which have a greatly altered in vitro tropism profile: in addition to having acquired the ability to infect the Mus dunni cell line, these stocks have lost the ability to infect certain other mouse cell lines, notably the Mus musculus domesticus wild mouse SC-1 cell line which is extraordinarily sensitive to MuLV infection. Strikingly, infection of neonatal mice with RF.AMLV or MoMuLV eco also results in the generation of similar ecotropic variants (i.e. acquired dunni-tropism) in vivo. In addition, mice of strains which express high levels of endogenous ecotropic MuLV demonstrate an age-dependent increase in the percentage of these ecotropic variants recoverable from various tissues. Studies indicate that 1) pseudotyping of the ecotropic MuLV mink cell focus-inducing (MCF) MuLV is not responsible for the observed change in infective host range; 2) no major recombination event detectable by restriction endonuclease mapping or differential hybridization is involved; and 3) significant differences in the latency and histopathology of the disease induced by various RF.AMLV stocks appears to correlate with the percentage of different ecotropic variants within the viral population.